A membrane form of brain L-glutamate decarboxylase: Identification, isolation, and its relation to insulin-dependent diabetes mellitus

A membrane form of L-glutamate decarboxylase (GAD) was identified and purified to apparent homogeneity from hog brain. The purified GAD was established as an integral membrane protein by phase-partitioning assay, charge-shift electrophoresis, and chromatography on a hydrophobic interaction column. This membrane GAD has a native molecular mass of 96 ± 5 kDa and is a homodimer of 48 ± 3-kDa subunits. Immunoprecipitation and immunoblotting tests revealed the presence of antibodies against this membrane GAD in sera from patients with insulin-dependent diabetes meilitus. Since this form of GAD appears to be an integral membrane protein and is presumed to have extracellular domains exposed, it seems reasonable to suggest that membrane GAD is more likely than soluble GAD to be involved in the pathogenesis of insulin-dependent diabetes and related autoimmune disorders such as stiff-man syndrome. y-Aminobutyric acid (GABA) is the major inhibitory neurotransmitter in the mammalian central nervous system (1) and also serves signaling (2) and trophic (3) roles in several neuronal and non-neuronal tissues. The rate-limiting step in GABA biosynthesis is the decarboxylation of L-glutamate by L-glutamate decarboxylase (GAD; EC 4.1.1.15). GAD has been implicated in several neuronal disorders, such as epilepsy (4), schizophrenia (5), and stiff-man syndrome (6), and has been identified as the autoantigen in insulin-dependent diabetes mellitus (IDDM) (7) and stiff-man syndrome (8). Several forms of soluble GAD (sGAD), notably GAD-65 and GAD-67 (65 and 67 kDa, respectively), have been extensively studied, ranging from kinetic studies, to studies of regional and cellular/subcellular distribution, to structural analysis (for review, see ref. 9). From sequence information, it is clear that none of the sGADs, including GAD-65 and GAD-67, contains a stretch of hydrophobic amino acids long enough to span the membrane (about 20 residues), a typical feature for integral membrane proteins, or contains the appropriate consensus sequences for the attachment of GAD to membranes through fatty acylation via esterification, N-myristoylation, or glypiation (9-11). Unlike sGAD, little is known with certainty about the structure and function ofmembraneGAD (mGAD) despite the fact that about 50%6 of the total GAD activity in the brain is attributable to mGAD (12, 13). GAD can interact with membranes by ionic or hydrophobic mechanisms. It was reported that GAD could become associated with membranes in the presence of Ca2+ (14). Covarrubias and Tapia (15, 16) showed that this Ca2+-induced binding of GAD to the membranes occurred predominantly with pyridoxal 5'-phosphate (PLP)-dependent GAD, suggesting that mGAD might be difThe publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact. ferent from sGAD in terms ofaffinity toward its cofactor, PLP. Martin and Martin (17) have shown that apoGAD has a strong affinity for polyanions (e.g., hexasulfate) and hence is attracted to synaptic vesicles, whose cytoplasmic face is known to be enriched in acidic phospholipids (10). However, Chang and Gottlieb (13) have demonstrated that 60% of GAD is membrane-bound and can be released only with detergent (0.2% Triton X-100) but not-with high salt (1 M NaCl or KCl), suggesting that the interaction between GAD and membranes is predominantly hydrophobic. Christgau et al. (18) reported that pancreatic (3 cells expressed two autoantigenic forms of GAD, GAD-65 and GAD-64. GAD-65 is a soluble protein, whereas GAD-64 is firmly membrane anchored and can be released only by detergent, such as Triton X-100, but not by any of the agents known to release peripheral membrane proteins. Christgau et al. (19) have shown that the autoantigen GAD-65 (previously referred to as GAD-64 by the same authors) is anchored to membranes of microvesicles in 3-cells by palmitoylation in the N-terminal domain. Reetz et aL (20) also reported colocalization ofGAD and GABA with synapticlike microvesicles in (3 cells and (3cell lines. Recently Solimena et al. (21) reported association ofGAD-65 but not GAD-67 with the Golgi complex of transfected Chinese hamster ovary cells. Further, the same authors showed that the association of GAD-65 with the Golgi complex was mediated by the N-terminal region. Hence, although GAD-65 is not an integral protein, it can be associated with membranes through posttranslational modification. In regard to possible functions, it has been proposed that mGAD may be involved in processing of memory and information and regulation of the central nervous system (22, 23). However, no definite results have been reported yet to support the above propositions. Thus it is clear that the information regarding the structure and function ofmGAD is quite limited. Therefore, we decided to undertake the purification of mGAD from porcine brain to homogeneity and to obtain more precise information about its structure, properties, and function. We present evidence indicating the presence of at least three different forms of mGAD, referred to as mGADI, -II, and -III. In addition, we describe the isolation, purification, and characterization of mGADI, which appears to be an integral membrane protein and is different from any other forms ofGAD that have been reported (9). Evidence suggesting that mGADI may be involved in certain autoimmune disorders such as IDDM is also included. mGADII is also an integral membrane protein and is an autoantigen for IDDM, whereas mGADIII is a peripheral membrane protein and its association with membranes is Abbreviations: GABA, y-aminobutync acid; GAD, L-glutamate decarboxylase; mGAD, membrane GAD; sGAD, soluble GAD; AET, S-(2-aminoethyl)isothiouronium bromide hydrobromide; PLP, pyridoxal 5'-phosphate; IDDM, insulin-dependent diabetes mellitus. iTo whom reprint requests should be addressed.